Also, this gives a basis for why large-scale decoders can learn shared embeddings across individuals (see our paper on POYO: poyo-brain.github.io). And expect more of this to come soon!